California refiners anticipate broad effects of possible state MTBE ban

California refiners are attempting to prepare for the unknown, with a possible ban on the gasoline additive MTBE pending in the state legislature. California's gasoline producers would need significant time and capital to modify their plants to produce gasoline that meets the state's stringent specifications without using MTBE. Shown here is Ultramar Diamond Shamrock Corp.'s Wilmington, Calif., refinery. Photo courtesy Ultramar Diamond Shamrock.

MTBE has been found in many water sources in California and other states. The contaminant imparts a noticeable odor and taste to drinking water.

Although relatively few examples of water well contamination have been identified, there is a movement in the California legislature to scale back or eliminate MTBE use in the state.

While other states have joined the bandwagon, California appears to be moving most rapidly in response to the perceived threat. Because the public is uncertain about the effects of low concentrations of MTBE in drinking water, regulators are under pressure to ban the material.

Some petroleum companies have taken the position that, even though MTBE does not pose any clear health risks, if consumers are concerned that such health risks might exist, refiners would prefer to supply some other type of fuel.

Refiners would meet California Air Resources Board (CARB) Phase II reformulated gasoline specifications through the use of ethanol, other oxygenates, or perhaps no oxygenate at all. In the event of an MTBE ban, however, time and capital would be needed to enable refiners to make needed refinery retrofits and to allow the pipeline infrastructure to respond.

MTBE contamination

Pipeline spills and leaks have contributed appreciable MTBE contamination. Two-stroke engines-common on small watercraft such as jet skis and boats with outboard engines-emit significant amounts of unburned fuel in their exhaust.

Studies in California have shown that, after busy summer weekends, MTBE is found in the water in lakes that are used both for recreation and as reservoirs for metropolitan use.

MTBE can impart an objectionable taste and odor to water at quite low concentrations. Some studies have shown that sensitive individuals can detect MTBE at concentrations below 40 ppb in water.

The California Department of Health Services has proposed a drinking water standard of 5 ppb, based on taste and odor thresholds. That standard is based not on known human health effects but rather on public perceptions.

The health effects of low-level MTBE exposure are unclear. The California legislature authorized a major study of the health impacts of MTBE and its alternatives. The study, performed by the University of California at Davis, did not encompass any original toxicological research but included an exhaustive review of the existing literature.

It is impossible to prove that no negative health effects can be associated with ingestion of a chemical like MTBE. Hence, it was quite unlikely at the outset that the UC Davis study would be able to offer concerned citizens confidence that MTBE is certainly safe.

Based on the evidence available, UC Davis concluded that MTBE is an animal carcinogen. That study recommends seeking alternatives to MTBE-such as substituting ethanol and producing gasoline meeting CARB specifications without the use of oxygenates-but it recognizes that these alternatives may have unknown health impacts of their own.

There are courses of action other than banning or restricting MTBE blending that would mitigate MTBE contamination.

When MTBE from gasoline is found in groundwater, the hydrocarbon constituents of the gasoline from which it came were released also; MTBE is only part of the contamination problem. Banning MTBE would remove only the MTBE part of the problem without addressing the hydrocarbon constituents.

Some voices in the debate favor more rigorous steps to reduce or prevent any gasoline contamination, whether or not MTBE is involved. Such steps might include: expanded underground storage tank inspection and maintenance, more rigorous pipeline testing, and phasing out or modifying two-stroke watercraft engines.

These alternatives cannot be dismissed out of hand, and the typical citizen probably would not be pleased with the thought of gasoline contamination of his or her drinking water, even in the absence of MTBE.

However, recent work in California suggests that a ban on MTBE can be instituted without incurring politically unacceptable cost increases, provided industry is given adequate time to respond. Consequently, there is a significant probability that, even if steps such as those listed previously are adopted, MTBE might still be banned in California.

Effects on refiners

The overlapping regulations make response to an MTBE ban more difficult.

Blending CARB gasoline requires simultaneous control of eight key variables, as shown in Table 1 [58,658 bytes].

CARB regulations provide for flat limits, as shown, but relatively little gasoline is produced to the flat-limit specifications. Cap limits, on the other hand, are maximums for each parameter for any given batch of gasoline.

Cap limits are applicable to all gasoline produced in California, whether it is made under the averaging system or the CARB predictive model.

Averaging limits are the average specification a refiner must meet over time to take advantage of the averaging part of the CARB gasoline program. The predictive model uses an algorithm to estimate emissions based on quality parameters.

In fact, almost all gasoline in California is certified with the predictive model.

The last column in the table shows typical test results for summertime gasoline marketed in California.

California refiners use the predictive model to stretch the abilities of their refineries to make CARB-compliant gasoline. Refiners find it to their advantage to reduce sulfur, benzene, and olefins to quite low levels, in search of more room on the 90% and 50% distillation points (T90 and T50). Maximizing those parameters allows refiners to keep the maximum volume of gasoline in the pool.

EPA regulations are generally consistent with CARB regulations, except in the area of oxygen content. EPA mandates a minimum oxygen content of 1.8 wt %, which corresponds to an MTBE concentration of about 11 vol %. CARB does not require any minimum oxygen content in summer gasoline and imposes a 2.7 wt % limit on oxygen.

MTBE is very important to CARB gasoline for a variety of reasons. If MTBE is removed from gasoline, other offsetting changes will have to be made.

The principal influences of MTBE on gasoline blending are based on dilution, octane, and oxygen.

The diluent effects of MTBE are important. When MTBE is added to a gasoline blend, the concentrations of objectionable compounds such as benzene, aromatics, or sulfur-bearing compounds are proportionately reduced. Removing MTBE or reducing its concentration would increase the concentration of those objectionable compounds unless some other step is taken to offset the loss of dilution.

Likely alternatives to MTBE-particularly ethanol or the "no-oxygenate" alternative-have less dilution potential than MTBE.

The octane effects of MTBE are appreciable and important.

Refiners have responded to aromatics and benzene limits by reducing the volume and severity of naphtha reforming, thus reducing the level of octane available from this important source. Limits on sulfur and olefins are achieved in part by hydrotreating gasoline produced by the fluid catalytic cracking (FCC) unit. But hydrotreating FCC gasoline is accompanied by some loss of octane.

Blending MTBE into California gasoline has mitigated the negative octane consequences of changes in other important processing steps.

The relationship of emissions to oxygen content, as described in the CARB predictive model and the EPA complex model, is not straightforward, and the two models view the influence of oxygen differently.

The predictive model contemplates oxygen content no higher than 2.8 wt %, and it views high oxygen content as being associated with increased NOx emissions. The complex model contemplates oxygen at higher ranges but not below 1.8 wt %.

The complex model does not assign such negative NOx consequences to high oxygen levels as does the predictive model. Blending 10 vol % ethanol into gasoline results in a blend with about 3.5 wt % oxygen, which is outside the range of CARB gasoline. At such levels, the predictive model requires extreme reductions in other controlled parameters to keep calculated vehicle emissions in the acceptable zone.

Alternatives to MTBE

• Ethanol can be substituted for MTBE, albeit with considerable difficulty.
• Other oxygenates, such as ethyl tertiary butyl ether (ETBE), tertiary amyl methyl ether (TAME), di-isopropyl ether, or tertiary-butyl alcohol (TBA), could be used.
• RFG could be manufactured without any oxygenate, although some change in federal regulations would be needed.

Other oxygenates are possibly easier to use, but supply would be a problem. And these compounds have some unknowns of their own, from a toxicology point of view. Consequently, the other oxygenates are not viewed as a likely long-term solution.

Oxygenate-free RFG is a viable proposal and already is manufactured in limited volumes in California. The prospects for needed federal regulatory changes make this an uncertain proposition statewide, however.

Ethanol

Ethanol has not been used regularly as a component of California gasoline since the implementation of the CARB Phase II program. Ethanol's vapor pressure effects on gasoline are a drawback compared with MTBE, which has relatively low vapor pressure.

Since California lacks any state tax incentives for ethanol use, and large-volume sources are distant from the state, refiners have not been as attracted to ethanol in California as they have been in Midwestern states.

Ethanol poses some other drawbacks as a substitute for MTBE. First, using ethanol in lieu of MTBE requires the refiner to reduce the vapor pressure of the crude oil-derived gasoline components by about 1.3 psig, which is quite difficult when blending to a 7.0 psi Rvp limit. An Rvp waiver has been proposed to help mitigate this problem, but CARB has not proposed adoption of such a waiver, as it believes the waiver would increase emissions of volatile hydrocarbon compounds.

Second, if ethanol is blended into gasoline at the 10 vol % level typical of Midwestern gasohol blends, the oxygen content of the gasoline is about 3.5 wt %. CARB's predictive model does not anticipate such oxygen levels, which exceed current limits under CARB regulations.

The predictive model associates increased NOx emissions with such high oxygen levels. This would require a refiner to make severe reductions in other gasoline parameters to stay within allowable NOx performance characteristics.

Thus, while blending ethanol at the 10 vol % level would seem to achieve about the same dilution as refiners gain from MTBE, the predictive model raises the hurdle refiners must meet.

Finally, unlike MTBE or its other substitutes, ethanol requires blending at the truck-loading point. Shipment of ethanol blends in multi-fuel pipelines is prevented by potential water contamination problems. Consequently, current practices call for shipping the hydrocarbon component separately from the ethanol. The two are then blended as the product is loaded into trucks.

The infrastructure to support such blending in California does not exist. Many terminals once engaged in eth- anol blending have since been modified. It has been estimated that 2 years would be needed to install adequate ethanol blending facilities statewide to support a change from MTBE to ethanol.

Ethanol recently has been reintroduced as a California gasoline constituent, however. Tosco Corp. has been test-marketing ethanol blends manufactured in its San Francisco Bay area refineries. Distribution has been limited to service stations that are directly accessible to the refinery by truck.

Tosco recently has announced an extension and an expansion of the limited ethanol program, based on an agreeable experience and a favorable consumer response.

Other oxygenates

EPA-approved oxygenates include TAME, ETBE and TBA, among others. There is a small amount of TAME used currently in California gasoline.

Supply of these non-ethanol oxygenate alternatives is somewhat uncertain. None is produced in adequate volumes to replace the MTBE used in California.

Investment would be needed to increase supply, and it is unclear that adequate investment would be timely or ever materialize. And prospective investors might reasonably conclude that these oxygenates would carry the same sorts of regulatory risk now associated with MTBE.

Toxicology of these oxygenates has not been well-studied. As a group, these materials are thought to have characteristics similar in many respects to MTBE. Substituting one or more of these materials for MTBE is generally thought to carry unknown risks of future problems similar to those bedeviling MTBE.

The UC Davis study identified even less toxicological data on these materials than on MTBE, and no group has come forward in California strongly favoring any of these oxygenates as MTBE substitutes.

No-oxygenate alternative

The CARB RFG program, which applies statewide, does not require the use of any oxygenate in summertime gasoline. The predictive model under which CARB gasolines must be certified is compatible with such blends, although losing the dilution effect of oxygenates makes specifications for the crude oil-derived portion of the gasoline more difficult to meet.

CARB RFG containing no oxygen has been marketed in northern California areas not subject to the federal mandate. Tosco has announced retail marketing, in addition to its ethanol program, of limited amounts of CARB gasoline containing no oxygenates. Like the ethanol program, this gasoline is available only in limited areas accessible by truck from Tosco's San Francisco area refineries.

Other refiners have manufactured CARB gasoline containing little or no oxygen but have not established programs for marketing of these blends on a regular basis. It is important to note that, while it is relatively easy for refiners to make small test batches of such fuels, major capital expenditures would be necessary before a refiner could manufacture all its gasoline to meet such strict specifications.

The federal RFG program, which applies only in Southern California, requires minimum oxygen content in California as it does in other federal RFG areas. Two bills have been introduced in the U.S. Congress to simplify these regulations and allow the generally more stringent CARB program to supersede the federal program in California. If these bills are enacted, refiners will acquire another degree of freedom to meet clean fuels goals in Southern California.

Refinery modifications

The hydrocarbon component of gasoline made without MTBE would have to have characteristics superior to those that currently exist, because of the loss of dilution. The most likely areas for modifications are: alkylation unit expansions, FCC gas processing, depentanization of gasoline blend stocks, and further FCC gasoline fractionation.

Refiners have a variety of methods available to improve gasoline quality.

Gasoline sulfur, already extremely low in California at 20 ppm or less on average, might be further reduced by expanding FCC gasoline desulfurization, at some expense to octane quality.

To make room for ethanol blending, gasoline vapor pressure can be decreased by depentanizing more streams in summertime. Removing pentane would directionally increase T50, however, and some other fractionation changes would be needed to reduce T90 adequately to counterbalance the change.

Many of the changes that refiners would make in response to an MTBE ban have negative consequences of one kind or another. An exception is the use of more alkylate as gasoline blend stock, particularly propylene alkylate.

Alkylate

Alkylate is a key component of reformulated gasolines, particularly in California, where so many constituents are limited. Alkylate lacks benzene, other aromatics, olefins, and sulfur and has satisfactory octane and vapor pressure characteristics. It is an ideal CARB RFG blend stock.

Interestingly, propylene alkylate may become more important in the future because it has more favorable distillation characteristics for CARB RFG than does conventional butylene alkylate. The T50 and T90 for propylene alkylate are 30-40° F. lower than for butylene alkylate see Table 2 [30,926 bytes]. One possible refiner response to eliminating oxygen and losing the dilution effect of MTBE would be to reduce T50 and T90. Consequently, although propylene alkylate has less-favorable octane characteristics than butylene alkylate, these other properties may be of adequate importance to shift refiner preference to the lighter material.

The theoretical Rvp of debutanized propylene alkylate is slightly higher than butylene alkylate, although both are well below regulatory limits. Most alkylate in trade is not fully debutanized or includes appreciable pentane, and observed Rvp levels in commercially available alkylate are typically much higher than the theoretical levels shown in the table.

The data in the table indicate that it is possible to achieve highly desirable Rvp levels with alkylate via debutanization and possibly depentanization. And, if the alkylate market expands substantially due to an MTBE ban, there will be pressure placed on suppliers to reduce Rvp to levels consistent with manufacturing CARB gasoline.

Heightened attractiveness of propylene for alkylation runs counter to recent trends in the U.S. industry.

High propylene values for petrochemical applications have drawn propylene feeds from alkylation units east of the Rockies, and refinery-based polypropylene capacity is being installed even in California. Break-even values for propylene in alkylation on the U.S. Gulf Coast have been about 10¢/gal below petrochemical value recently, and alkylation economics were much worse in 1997.

Hence, an appreciable economic hurdle exists to attracting propylene back into alkylation units.

The demand for alkylate from external sources could increase very substantially if MTBE is banned in California. Over the long term, California refiners could respond to an MTBE ban by erecting new alkylation units or expanding existing plants, although, over the short to intermediate term, time for such steps would not be available.

Depending on how an MTBE ban is instituted, one solution would be for California refiners to seek alkylate imports of 50,000-100,000 b/d.

Timing and costs

Over the short term, an MTBE ban would be catastrophic to gasoline supplies. California refiners simply are not prepared to manufacture CARB gasoline without MTBE, and refiners outside the state would be unreliable suppliers of the volumes needed. The California Energy Commission (CEC) has estimated the shortfall at 15-40% of California demand, or perhaps 300,000 b/d.

While refiners on the U.S. Gulf Coast have appreciable ability to manufacture CARB gasoline, given appropriate price signals, delivering such volumes to California could be difficult due to limitations in the Jones Act tanker fleet. Furthermore, over the short term, pipeline terminals could not be made ready to blend ethanol, removing an important MTBE substitute from contention.

Without MTBE and unable to use ethanol, California refiners would be forced to rely on other oxygenates to meet federal requirements.

Over the intermediate term-perhaps 2-4 years-a ban could be implemented by relying on waterborne imports of appreciable fuel to supplement California production. During that time frame, ethanol is a viable substitute for MTBE, and distribution problems could be solved. Nevertheless, if California refiners are not given adequate time to design, permit, and install substantial modifications, external fuel sources will be needed.

Over the long term-perhaps 5 years or more-MTBE could be banned without triggering severe adverse price reactions in the state. This is the safest course, from the point of view of fuel supply reliability.

The practical availability of external sources of CARB gasoline or alkylate, and the price impact of delivering the material to the state, may limit the flexibility California regulators feel in the timing of an MTBE ban. The West Coast has been mostly a self-contained market for petroleum products. Importing such large quantities of alkylate, or finished CARB gasoline, would link California markets to world pricing in ways that are beyond the common experience of California consumers.

To make such large volumes of alkylate available to California, refiners in other parts of the world would have to make substantial adjustments to their own operations, adjustments that could not be made without adequate price incentive. There is potential for substantial upward price pressure, which could result in consumer backlash against an MTBE ban.

The costs of an MTBE ban are highly dependent on the pace of change. The CEC has adopted a report showing that, over the intermediate term, the average cost of producing gasoline in California would increase by 6-12¢/gal using either the ethanol or the no-oxygenate alternative. By waiting for the long term, thus giving refiners time to make modifications, cost increases could be held to no more than about 4¢/gal, and perhaps substantially less.

Outlook

The cost levels estimated in the CEC work are considered low enough that an MTBE ban cannot be ruled out. The CEC study implies that, at least in the long term, CARB fuel supplies would not be threatened.

Consequently the likelihood of an MTBE ban cannot be dismissed, and the issue was expected to be raised by the California legislature this month.

Lessons for the future

Hindsight is 20/20, but at the time of the Clean Air Act Amendments of 1990 and the development of the CARB RFG program, none of the groups now opposing MTBE identified water contamination as a problem of adequate scale to address. The water contamination problem, whatever its scale, is an unintended negative consequence of a well-intended program.

The history of gasoline quality regulation has always been punctuated by unintended, undesirable consequences. During the 1960s and 1970s, EPA focused on tetraethyl lead and reducing airborne lead contamination. In retrospect, this was probably a sound goal with worthwhile health consequences, but it lead to more severe reforming and increases in the concentrations of benzene and aromatics in gasoline.

By the late 1980s EPA recognized these contaminants as new problems, and the federal and California RFG programs were adopted to remedy this unintended consequence of lead phase-out. Now regulators have discovered that the newest air advances have contaminated the water. Overall, we are making progress, as benzene is probably a smaller health problem than lead was, and MTBE is a smaller problem than benzene was. Nevertheless, future programs probably will have unintended consequences of their own.

The solution is not more study, as reformulated gasoline was heavily studied. Rather, the solution is to remain flexible and recognize that fuel specifications will continue to evolve.

Astute refiners will recognize that no set of specifications is final.

The Author

John Vautrain is vice-president and director of Purvin & Gertz Inc. He manages Purvin & Gertz's Long Beach, Calif., office and has been consulting in U.S. West Coast and Asia-Pacific energy issues for 15 years. His consulting work includes analysis of environmental issues related to industrial and consumer fuels, as well as their production, trade, and marketing. Vautrain holds a bachelors degree in chemistry from the University of Texas and a masters degree in chemical engineering from the University of Utah. He is a member of the American Institute of Chemical Engineers, the Society of Petroleum Engineers, and the International Association for Energy Economics.

Copyright 1999 Oil & Gas Journal. All Rights Reserved.